import os
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
import seaborn as sns; sns.set()
from numpy import *
from scipy import stats
from pandas.plotting import scatter_matrix
import sklearn
import warnings

data = pd.read_excel(r'Attrition Data Exercise.xlsx')

data.columns

Index(['Age', 'Attrition', 'BusinessTravel', 'DailyRate', 'Department',
       'DistanceFromHome', 'Education', 'Educ_bucket', 'EducationField',
       'EmployeeCount', 'EmployeeNumber', 'EnvironmentSatisfaction',
       'EnvironmentSatisfaction_Bucket', 'Gender', 'HourlyRate',
       'JobInvolvement', 'JobInvolvement_bucket', 'JobLevel', 'JobRole',
       'JobSatisfaction', 'JobSatisfaction_bucket', 'MaritalStatus',
       'MonthlyIncome', 'MonthlyRate', 'NumCompaniesWorked', 'Over18',
       'OverTime', 'PercentSalaryHike', 'PerformanceRating',
       'RelationshipSatisfaction', 'RelationshipSatisfaction_bucket',
       'StandardHours', 'StockOptionLevel', 'TotalWorkingYears',
       'TrainingTimesLastYear', 'WorkLifeBalance', 'WorkLifeBalance_bucket',
       'Company Tenure (yrs)', 'Company Tenure (yrs)_Bucket',
       'YearsInCurrentRole', 'YearsSinceLastPromotion', 'YearsWithCurrManager',
       'Location'],
      dtype='object')

data.head(5)

	Age	Attrition	BusinessTravel	DailyRate	Department	DistanceFromHome	Education	Educ_bucket	EducationField	EmployeeCount	...	TotalWorkingYears	TrainingTimesLastYear	WorkLifeBalance	WorkLifeBalance_bucket	Company Tenure (yrs)	Company Tenure (yrs)_Bucket	YearsInCurrentRole	YearsSinceLastPromotion	YearsWithCurrManager	Location
0	41	Yes	Travel_Rarely	1102	Sales	1	2	College	Life Sciences	1	...	8	0	1	Bad	6	5 to 10 years	4	0	5	India
1	49	No	Travel_Frequently	279	Research & Development	8	1	Below College	Life Sciences	1	...	10	3	3	Better	10	10 to 15 yrs	7	1	7	China
2	37	Yes	Travel_Rarely	1373	Research & Development	2	2	College	Other	1	...	7	3	3	Better	0	Less than 2 years	0	0	0	India
3	33	No	Travel_Frequently	1392	Research & Development	3	4	Masters	Life Sciences	1	...	8	3	3	Better	8	5 to 10 years	7	3	0	India
4	27	No	Travel_Rarely	591	Research & Development	2	1	Below College	Medical	1	...	6	3	3	Better	2	3 to 5 yrs	2	2	2	India

5 rows × 43 columns

Data Preprocessing

Rounding of some of the columns to make easier and larger buckets for better analysis

data.DailyRate = data.DailyRate.round(-2)
data.HourlyRate = data.HourlyRate.round(-1)
data.MonthlyIncome = data.MonthlyIncome.round(-3)
data.MonthlyRate = data.MonthlyRate.round(-3)
data.Age = data.Age.round(-1)

data_original = data

Replace Yes and No with 1 and 0 for attrition variable

Num_val = {'Yes':1, 'No':0}
data['Attrition'] = data["Attrition"].apply(lambda x: Num_val[x])

data.iloc[:, 30 : 40]

	RelationshipSatisfaction_bucket	StandardHours	StockOptionLevel	TotalWorkingYears	TrainingTimesLastYear	WorkLifeBalance	WorkLifeBalance_bucket	Company Tenure (yrs)	Company Tenure (yrs)_Bucket	YearsInCurrentRole
0	Low	80	0	8	0	1	Bad	6	5 to 10 years	4
1	Very high	80	1	10	3	3	Better	10	10 to 15 yrs	7
2	Medium	80	0	7	3	3	Better	0	Less than 2 years	0
3	High	80	0	8	3	3	Better	8	5 to 10 years	7
4	Very high	80	1	6	3	3	Better	2	3 to 5 yrs	2
...	...	...	...	...	...	...	...	...	...	...
1465	High	80	1	17	3	3	Better	5	5 to 10 years	2
1466	Low	80	1	9	5	3	Better	7	5 to 10 years	7
1467	Medium	80	1	6	0	3	Better	6	5 to 10 years	2
1468	Very high	80	0	17	3	2	Good	9	5 to 10 years	6
1469	Low	80	0	6	3	4	Best	4	3 to 5 yrs	3

1470 rows × 10 columns

Getting stats of the features. Droppping features with 0 variance, features which are irrelevant like Employee Number and Dropping Categorical Columns which already have Numerical Ordered Values which can be used like Educ_bucket

data.describe()

	Age	Attrition	DailyRate	DistanceFromHome	Education	EmployeeCount	EmployeeNumber	EnvironmentSatisfaction	HourlyRate	JobInvolvement	...	RelationshipSatisfaction	StandardHours	StockOptionLevel	TotalWorkingYears	TrainingTimesLastYear	WorkLifeBalance	Company Tenure (yrs)	YearsInCurrentRole	YearsSinceLastPromotion	YearsWithCurrManager
count	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000	1470.0	1470.000000	1470.000000	1470.000000	1470.000000	...	1470.000000	1470.0	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000
mean	37.061224	0.161224	801.836735	9.192517	2.912925	1.0	1024.865306	2.721769	65.986395	2.729932	...	2.712245	80.0	0.793878	11.279592	2.799320	2.761224	7.008163	4.229252	2.187755	4.123129
std	9.667885	0.367863	406.000802	8.106864	1.024165	0.0	602.024335	1.093082	20.796262	0.711561	...	1.081209	0.0	0.852077	7.780782	1.289271	0.706476	6.126525	3.623137	3.222430	3.568136
min	20.000000	0.000000	100.000000	1.000000	1.000000	1.0	1.000000	1.000000	30.000000	1.000000	...	1.000000	80.0	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000
25%	30.000000	0.000000	500.000000	2.000000	2.000000	1.0	491.250000	2.000000	50.000000	2.000000	...	2.000000	80.0	0.000000	6.000000	2.000000	2.000000	3.000000	2.000000	0.000000	2.000000
50%	40.000000	0.000000	800.000000	7.000000	3.000000	1.0	1020.500000	3.000000	70.000000	3.000000	...	3.000000	80.0	1.000000	10.000000	3.000000	3.000000	5.000000	3.000000	1.000000	3.000000
75%	40.000000	0.000000	1200.000000	14.000000	4.000000	1.0	1555.750000	4.000000	80.000000	3.000000	...	4.000000	80.0	1.000000	15.000000	3.000000	3.000000	9.000000	7.000000	3.000000	7.000000
max	60.000000	1.000000	1500.000000	29.000000	5.000000	1.0	2068.000000	4.000000	100.000000	4.000000	...	4.000000	80.0	3.000000	40.000000	6.000000	4.000000	40.000000	18.000000	15.000000	17.000000

8 rows × 27 columns

threshold = 0.0

data = data.drop(data.std()[data.std() == threshold].index.values, axis=1)
data = data.drop('EmployeeNumber', axis = 1)
data = data.drop(['Educ_bucket', 'EnvironmentSatisfaction_Bucket', 'JobInvolvement_bucket', 'JobSatisfaction_bucket', 'RelationshipSatisfaction_bucket', 'WorkLifeBalance_bucket', 'Company Tenure (yrs)_Bucket'], axis = 1)

data_new = data

Correlation Matrix : Could have removed highly correlated variables if model performance had to be improved more

fig, ax=plt.subplots(figsize=(200,200))
plt.title('Correaltion Plot',fontsize=50)
sns.heatmap(data.corr(), mask=np.zeros_like(data.corr(), dtype=np.bool), cmap=sns.diverging_palette(220, 10, as_cmap=True),
            square=True, ax=ax, annot=True,linewidths=3)
plt.xticks(rotation=90, fontsize=100)
plt.yticks(rotation=0, fontsize=100)
plt.show()

Getting target variable in Y and rest of the features in X

Y = data.Attrition

Y.shape

(1470,)

X = data.drop('Attrition', axis = 1)

X.iloc[:, 20 : 50]

	PercentSalaryHike	PerformanceRating	RelationshipSatisfaction	StockOptionLevel	TotalWorkingYears	TrainingTimesLastYear	WorkLifeBalance	Company Tenure (yrs)	YearsInCurrentRole	YearsSinceLastPromotion	YearsWithCurrManager	Location
0	11	3	1	0	8	0	1	6	4	0	5	India
1	23	4	4	1	10	3	3	10	7	1	7	China
2	15	3	2	0	7	3	3	0	0	0	0	India
3	11	3	3	0	8	3	3	8	7	3	0	India
4	12	3	4	1	6	3	3	2	2	2	2	India
...	...	...	...	...	...	...	...	...	...	...	...	...
1465	17	3	3	1	17	3	3	5	2	0	3	USA
1466	15	3	1	1	9	5	3	7	7	1	7	USA
1467	20	4	2	1	6	0	3	6	2	0	3	USA
1468	14	3	4	0	17	3	2	9	6	0	8	USA
1469	12	3	1	0	6	3	4	4	3	1	2	USA

1470 rows × 12 columns

Using OneHotEncoding for non ordered Categorical Features, OrdinalEncoding for ordered Categorical Features and StandardScaler to scale all numerical features

from sklearn.preprocessing import OneHotEncoder
ohe = OneHotEncoder()
from sklearn.preprocessing import OrdinalEncoder
oe = OrdinalEncoder()
from sklearn.compose import make_column_transformer
from sklearn.preprocessing import StandardScaler

X.iloc[:, 0 : 20]

	Age	BusinessTravel	DailyRate	Department	DistanceFromHome	Education	EducationField	EnvironmentSatisfaction	Gender	HourlyRate	JobInvolvement	JobLevel	JobRole	JobSatisfaction	MaritalStatus	MonthlyIncome	MonthlyRate	NumCompaniesWorked	Over18	OverTime
0	40	Travel_Rarely	1100	Sales	1	2	Life Sciences	2	Female	90	3	2	Sales Executive	4	Single	6000	19000	8	Y	Yes
1	50	Travel_Frequently	300	Research & Development	8	1	Life Sciences	3	Male	60	2	2	Research Scientist	2	Married	5000	25000	1	Y	No
2	40	Travel_Rarely	1400	Research & Development	2	2	Other	4	Male	90	2	1	Laboratory Technician	3	Single	2000	2000	6	Y	Yes
3	30	Travel_Frequently	1400	Research & Development	3	4	Life Sciences	4	Female	60	3	1	Research Scientist	3	Married	3000	23000	1	Y	Yes
4	30	Travel_Rarely	600	Research & Development	2	1	Medical	1	Male	40	3	1	Laboratory Technician	2	Married	3000	17000	9	Y	No
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1465	40	Travel_Frequently	900	Research & Development	23	2	Medical	3	Male	40	4	2	Laboratory Technician	4	Married	3000	12000	4	Y	No
1466	40	Travel_Rarely	600	Research & Development	6	1	Medical	4	Male	40	2	3	Healthcare Representative	1	Married	10000	21000	4	Y	No
1467	30	Travel_Rarely	200	Research & Development	4	3	Life Sciences	2	Male	90	4	2	Manufacturing Director	2	Married	6000	5000	1	Y	Yes
1468	50	Travel_Frequently	1000	Sales	2	3	Medical	4	Male	60	2	2	Sales Executive	2	Married	5000	13000	2	Y	No
1469	30	Travel_Rarely	600	Research & Development	8	3	Medical	2	Male	80	4	2	Laboratory Technician	3	Married	4000	10000	2	Y	No

1470 rows × 20 columns

numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns

numerical_ix

Index(['Age', 'DailyRate', 'DistanceFromHome', 'Education',
       'EnvironmentSatisfaction', 'HourlyRate', 'JobInvolvement', 'JobLevel',
       'JobSatisfaction', 'MonthlyIncome', 'MonthlyRate', 'NumCompaniesWorked',
       'PercentSalaryHike', 'PerformanceRating', 'RelationshipSatisfaction',
       'StockOptionLevel', 'TotalWorkingYears', 'TrainingTimesLastYear',
       'WorkLifeBalance', 'Company Tenure (yrs)', 'YearsInCurrentRole',
       'YearsSinceLastPromotion', 'YearsWithCurrManager'],
      dtype='object')

column_trans = make_column_transformer(
(OneHotEncoder(), ['Department', 'EducationField', 'Gender', 'JobRole', 'MaritalStatus', 'Over18', 'OverTime', 'Location']),
(OrdinalEncoder(), ['BusinessTravel']),
(StandardScaler(), ['Age', 'DailyRate', 'DistanceFromHome', 'Education',
       'EnvironmentSatisfaction', 'HourlyRate',
       'JobInvolvement', 'JobLevel', 'JobSatisfaction', 'MonthlyIncome',
       'MonthlyRate', 'NumCompaniesWorked', 'PercentSalaryHike',
       'PerformanceRating', 'RelationshipSatisfaction',
       'StockOptionLevel', 'TotalWorkingYears', 'TrainingTimesLastYear',
       'WorkLifeBalance', 'Company Tenure (yrs)', 'YearsInCurrentRole',
       'YearsSinceLastPromotion', 'YearsWithCurrManager']),    
remainder = 'passthrough')

X_new  = column_trans.fit_transform(X)

column_trans.get_feature_names

<bound method ColumnTransformer.get_feature_names of ColumnTransformer(remainder='passthrough',
                  transformers=[('onehotencoder', OneHotEncoder(),
                                 ['Department', 'EducationField', 'Gender',
                                  'JobRole', 'MaritalStatus', 'Over18',
                                  'OverTime', 'Location']),
                                ('ordinalencoder', OrdinalEncoder(),
                                 ['BusinessTravel']),
                                ('standardscaler', StandardScaler(),
                                 ['Age', 'DailyRate', 'DistanceFromHome',
                                  'Education', 'EnvironmentSatisfacti...
                                  'HourlyRate', 'JobInvolvement', 'JobLevel',
                                  'JobSatisfaction', 'MonthlyIncome',
                                  'MonthlyRate', 'NumCompaniesWorked',
                                  'PercentSalaryHike', 'PerformanceRating',
                                  'RelationshipSatisfaction',
                                  'StockOptionLevel', 'TotalWorkingYears',
                                  'TrainingTimesLastYear', 'WorkLifeBalance',
                                  'Company Tenure (yrs)', 'YearsInCurrentRole',
                                  'YearsSinceLastPromotion',
                                  'YearsWithCurrManager'])])>

def get_feature_names(column_transformer):
    """Get feature names from all transformers.
    Returns
    -------
    feature_names : list of strings
        Names of the features produced by transform.
    """
    # Remove the internal helper function
    #check_is_fitted(column_transformer)
    
    # Turn loopkup into function for better handling with pipeline later
    def get_names(trans):
        # >> Original get_feature_names() method
        if trans == 'drop' or (
                hasattr(column, '__len__') and not len(column)):
            return []
        if trans == 'passthrough':
            if hasattr(column_transformer, '_df_columns'):
                if ((not isinstance(column, slice))
                        and all(isinstance(col, str) for col in column)):
                    return column
                else:
                    return column_transformer._df_columns[column]
            else:
                indices = np.arange(column_transformer._n_features)
                return ['x%d' % i for i in indices[column]]
        if not hasattr(trans, 'get_feature_names'):
        # >>> Change: Return input column names if no method avaiable
            # Turn error into a warning
            warnings.warn("Transformer %s (type %s) does not "
                                 "provide get_feature_names. "
                                 "Will return input column names if available"
                                 % (str(name), type(trans).__name__))
            # For transformers without a get_features_names method, use the input
            # names to the column transformer
            if column is None:
                return []
            else:
                return [name + "__" + f for f in column]

        return [name + "__" + f for f in trans.get_feature_names()]
    
    ### Start of processing
    feature_names = []
    
    # Allow transformers to be pipelines. Pipeline steps are named differently, so preprocessing is needed
    if type(column_transformer) == sklearn.pipeline.Pipeline:
        l_transformers = [(name, trans, None, None) for step, name, trans in column_transformer._iter()]
    else:
        # For column transformers, follow the original method
        l_transformers = list(column_transformer._iter(fitted=True))
    
    
    for name, trans, column, _ in l_transformers: 
        if type(trans) == sklearn.pipeline.Pipeline:
            # Recursive call on pipeline
            _names = get_feature_names(trans)
            # if pipeline has no transformer that returns names
            if len(_names)==0:
                _names = [name + "__" + f for f in column]
            feature_names.extend(_names)
        else:
            feature_names.extend(get_names(trans))
    
    return feature_names

feat_names = get_feature_names(column_trans)

<ipython-input-889-aeb52d502674>:30: UserWarning: Transformer ordinalencoder (type OrdinalEncoder) does not provide get_feature_names. Will return input column names if available
  warnings.warn("Transformer %s (type %s) does not "
<ipython-input-889-aeb52d502674>:30: UserWarning: Transformer standardscaler (type StandardScaler) does not provide get_feature_names. Will return input column names if available
  warnings.warn("Transformer %s (type %s) does not "

X_new.shape

(1470, 53)

X_new

array([[ 0.        ,  0.        ,  1.        , ..., -0.0632959 ,
        -0.67914568,  0.24583399],
       [ 0.        ,  1.        ,  0.        , ...,  0.76499762,
        -0.36871529,  0.80654148],
       [ 0.        ,  1.        ,  0.        , ..., -1.16768726,
        -0.67914568, -1.15593471],
       ...,
       [ 0.        ,  1.        ,  0.        , ..., -0.61549158,
        -0.67914568, -0.31487349],
       [ 0.        ,  0.        ,  1.        , ...,  0.48889978,
        -0.67914568,  1.08689522],
       [ 0.        ,  1.        ,  0.        , ..., -0.33939374,
        -0.36871529, -0.59522723]])

Modelling

Splitting into Train and Test Sets

from sklearn.model_selection import train_test_split
import sklearn.model_selection as model_selection
X_train,X_test,Y_train,Y_test=train_test_split(X_new,Y,test_size=0.20,random_state=200)

X_train.shape

(1176, 53)

X_test.shape

(294, 53)

Y_train.shape

(1176,)

Y_test.shape

(294,)

Running Logistic Regression Model for benchmark

from sklearn.linear_model import LogisticRegression
logreg = LogisticRegression(C=1e5, penalty='l2', max_iter = 500)
logreg.fit(X_train, Y_train)
Y_pred = logreg.predict(X_test)
from sklearn import metrics
print("Accuracy:",metrics.accuracy_score(Y_test, Y_pred))

Accuracy: 0.8741496598639455

Running Random Forest Model with best Hyperparameters

Using Grid Search CV for Hyperparameter Tuning and Stratified Cross Validation for better sampling

from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import GridSearchCV
param_grid = {
    'n_estimators': [100, 250, 500],
    'max_features': ['auto', 'sqrt', 'log2'],
    'max_depth' : [5,6,7,8,9,10],
    'criterion' :['gini', 'entropy']
}
,
rf = RandomForestClassifier()
grid_search = GridSearchCV(estimator = rf, param_grid = param_grid, 
                          cv = 9, n_jobs = -1, verbose = 2)



grid_search.fit(X_train, Y_train)
grid_search.best_params_

Fitting 9 folds for each of 108 candidates, totalling 972 fits

[Parallel(n_jobs=-1)]: Using backend LokyBackend with 16 concurrent workers.
[Parallel(n_jobs=-1)]: Done   9 tasks      | elapsed:    3.0s
[Parallel(n_jobs=-1)]: Done 130 tasks      | elapsed:   10.8s
[Parallel(n_jobs=-1)]: Done 333 tasks      | elapsed:   24.8s
[Parallel(n_jobs=-1)]: Done 616 tasks      | elapsed:   45.3s
[Parallel(n_jobs=-1)]: Done 972 out of 972 | elapsed:  1.2min finished

{'criterion': 'gini',
 'max_depth': 10,
 'max_features': 'auto',
 'n_estimators': 250}

model = grid_search.best_estimator_
Y_pred = model.predict(X_test)
from sklearn import metrics
print("Accuracy:",metrics.accuracy_score(Y_test, Y_pred))

Accuracy: 0.8741496598639455

Checking for Recall for the model as we are most concerned with not classifying churners as non churners

from sklearn.metrics import classification_report,confusion_matrix
cr = classification_report(Y_test,Y_pred)
print(cr)

              precision    recall  f1-score   support

           0       0.88      0.99      0.93       253
           1       0.75      0.15      0.24        41

    accuracy                           0.87       294
   macro avg       0.81      0.57      0.59       294
weighted avg       0.86      0.87      0.84       294

model

RandomForestClassifier(max_depth=10, n_estimators=250)

model.feature_importances_

array([0.00276199, 0.00872382, 0.00971095, 0.00318348, 0.00749233,
       0.00745929, 0.00971985, 0.0023605 , 0.00796265, 0.0069103 ,
       0.0076735 , 0.00227082, 0.00349829, 0.01162021, 0.00146143,
       0.00240517, 0.00063344, 0.0073828 , 0.00670315, 0.00952692,
       0.00736391, 0.0085661 , 0.01809692, 0.        , 0.04029849,
       0.03971613, 0.00635151, 0.00818799, 0.00772115, 0.01294218,
       0.03325798, 0.03556215, 0.04849996, 0.02047774, 0.03176444,
       0.03115271, 0.0221354 , 0.02298059, 0.02997191, 0.04558261,
       0.03918754, 0.03199632, 0.0325047 , 0.00610841, 0.02206076,
       0.03054476, 0.06272628, 0.02560096, 0.02441481, 0.04889998,
       0.03100731, 0.02487155, 0.02998584])

feat_names

['onehotencoder__x0_Human Resources',
 'onehotencoder__x0_Research & Development',
 'onehotencoder__x0_Sales',
 'onehotencoder__x1_Human Resources',
 'onehotencoder__x1_Life Sciences',
 'onehotencoder__x1_Marketing',
 'onehotencoder__x1_Medical',
 'onehotencoder__x1_Other',
 'onehotencoder__x1_Technical Degree',
 'onehotencoder__x2_Female',
 'onehotencoder__x2_Male',
 'onehotencoder__x3_Healthcare Representative',
 'onehotencoder__x3_Human Resources',
 'onehotencoder__x3_Laboratory Technician',
 'onehotencoder__x3_Manager',
 'onehotencoder__x3_Manufacturing Director',
 'onehotencoder__x3_Research Director',
 'onehotencoder__x3_Research Scientist',
 'onehotencoder__x3_Sales Executive',
 'onehotencoder__x3_Sales Representative',
 'onehotencoder__x4_Divorced',
 'onehotencoder__x4_Married',
 'onehotencoder__x4_Single',
 'onehotencoder__x5_Y',
 'onehotencoder__x6_No',
 'onehotencoder__x6_Yes',
 'onehotencoder__x7_China',
 'onehotencoder__x7_India',
 'onehotencoder__x7_USA',
 'ordinalencoder__BusinessTravel',
 'standardscaler__Age',
 'standardscaler__DailyRate',
 'standardscaler__DistanceFromHome',
 'standardscaler__Education',
 'standardscaler__EnvironmentSatisfaction',
 'standardscaler__HourlyRate',
 'standardscaler__JobInvolvement',
 'standardscaler__JobLevel',
 'standardscaler__JobSatisfaction',
 'standardscaler__MonthlyIncome',
 'standardscaler__MonthlyRate',
 'standardscaler__NumCompaniesWorked',
 'standardscaler__PercentSalaryHike',
 'standardscaler__PerformanceRating',
 'standardscaler__RelationshipSatisfaction',
 'standardscaler__StockOptionLevel',
 'standardscaler__TotalWorkingYears',
 'standardscaler__TrainingTimesLastYear',
 'standardscaler__WorkLifeBalance',
 'standardscaler__Company Tenure (yrs)',
 'standardscaler__YearsInCurrentRole',
 'standardscaler__YearsSinceLastPromotion',
 'standardscaler__YearsWithCurrManager']

Creating dictionary of features and their importance to find out top 3 contributing features

dictionary = dict(zip(feat_names, model.feature_importances_))

my_keys = sorted(dictionary, key=dictionary.get, reverse=True)[:3]

my_keys

['standardscaler__TotalWorkingYears',
 'standardscaler__Company Tenure (yrs)',
 'standardscaler__DistanceFromHome']

Data Analysis

data_new.columns

Index(['Age', 'Attrition', 'BusinessTravel', 'DailyRate', 'Department',
       'DistanceFromHome', 'Education', 'EducationField',
       'EnvironmentSatisfaction', 'Gender', 'HourlyRate', 'JobInvolvement',
       'JobLevel', 'JobRole', 'JobSatisfaction', 'MaritalStatus',
       'MonthlyIncome', 'MonthlyRate', 'NumCompaniesWorked', 'Over18',
       'OverTime', 'PercentSalaryHike', 'PerformanceRating',
       'RelationshipSatisfaction', 'StockOptionLevel', 'TotalWorkingYears',
       'TrainingTimesLastYear', 'WorkLifeBalance', 'Company Tenure (yrs)',
       'YearsInCurrentRole', 'YearsSinceLastPromotion', 'YearsWithCurrManager',
       'Location'],
      dtype='object')

features1 = ['Age', 'Attrition', 'BusinessTravel', 'DailyRate', 'Department']
features2 = ['DistanceFromHome', 'Educ_bucket', 'EducationField'] 
features3 = ['WorkLifeBalance_bucket', 'Company Tenure (yrs)', 'YearsInCurrentRole', 'YearsSinceLastPromotion']
features4 = [ 'JobLevel', 'JobRole', 'JobSatisfaction_bucket', 'MaritalStatus']
features5 = ['MonthlyIncome', 'MonthlyRate', 'NumCompaniesWorked', 'Over18']
features6 = ['OverTime', 'PercentSalaryHike', 'PerformanceRating'] 
features7 = ['RelationshipSatisfaction_bucket', 'StockOptionLevel', 'TotalWorkingYears']
features8 = ['TrainingTimesLastYear', 'WorkLifeBalance_bucket', 'Company Tenure (yrs)_Bucket']
features9 = ['YearsInCurrentRole', 'YearsSinceLastPromotion', 'YearsWithCurrManager', 'Location']
features10 = ['EnvironmentSatisfaction_Bucket', 'JobInvolvement_bucket']

data_new

	Age	Attrition	BusinessTravel	DailyRate	Department	DistanceFromHome	Education	EducationField	EnvironmentSatisfaction	Gender	...	RelationshipSatisfaction	StockOptionLevel	TotalWorkingYears	TrainingTimesLastYear	WorkLifeBalance	Company Tenure (yrs)	YearsInCurrentRole	YearsSinceLastPromotion	YearsWithCurrManager	Location
0	40	1	Travel_Rarely	1100	Sales	1	2	Life Sciences	2	Female	...	1	0	8	0	1	6	4	0	5	India
1	50	0	Travel_Frequently	300	Research & Development	8	1	Life Sciences	3	Male	...	4	1	10	3	3	10	7	1	7	China
2	40	1	Travel_Rarely	1400	Research & Development	2	2	Other	4	Male	...	2	0	7	3	3	0	0	0	0	India
3	30	0	Travel_Frequently	1400	Research & Development	3	4	Life Sciences	4	Female	...	3	0	8	3	3	8	7	3	0	India
4	30	0	Travel_Rarely	600	Research & Development	2	1	Medical	1	Male	...	4	1	6	3	3	2	2	2	2	India
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1465	40	0	Travel_Frequently	900	Research & Development	23	2	Medical	3	Male	...	3	1	17	3	3	5	2	0	3	USA
1466	40	0	Travel_Rarely	600	Research & Development	6	1	Medical	4	Male	...	1	1	9	5	3	7	7	1	7	USA
1467	30	0	Travel_Rarely	200	Research & Development	4	3	Life Sciences	2	Male	...	2	1	6	0	3	6	2	0	3	USA
1468	50	0	Travel_Frequently	1000	Sales	2	3	Medical	4	Male	...	4	0	17	3	2	9	6	0	8	USA
1469	30	0	Travel_Rarely	600	Research & Development	8	3	Medical	2	Male	...	1	0	6	3	4	4	3	1	2	USA

1470 rows × 33 columns

Grouping by 'Attrition' column and seeing how behavior changes with different values for different features

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features1):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features2):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features3):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features4):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features5):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features6):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features7):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features8):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features9):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)

fig=plt.subplots(figsize=(50,50))
for i, j in enumerate(features10):
    plt.subplot(3, 2, i+1)
    plt.subplots_adjust(hspace = 1.0)
    sns.countplot(x=j,data = data_original, hue='Attrition')
    plt.xticks(rotation=90, fontsize = 30)
    plt.yticks(rotation=00, fontsize = 30)
    plt.xlabel(j, fontsize=40)
    plt.ylabel('count', fontsize=40)
    plt.title("No. of employee", fontsize=40)